Research On Application Of Stochastic Resonance Theory In Spectrum Sensing

In recent years, with the development of wireless business, wireless spectrum resourcesgradually become less. Meanwhile, the disadvantages of fixed spectrum allocation continue toappear. How to improve the spectrum utilization becomes a very concerned technical problemamong people. Total available spectrum resources of network and users can be dynamicallyincreased through cognitive radio technology and the cognitive radio technology is able to provide apossible solution for spectrum allocation and provide greater flexibility for the network andterminal. As a consequence, cognitive radio has a great significance for the research on wirelesscommunication. Spectrum sensing is not only the most basic but also the most critical technology ofcognitive radio. Currently, one of the biggest challenges of spectrum sensing is how to detect weaksignals.Because the performance of existing spectrum sensing algorithm while detecting low SNRsignal is not ideal and detection probability is low, this paper proposes a spectrum sensing methodthrough the best stochastic resonance theory. This algorithm introduces the principle of stochasticresonance in physics and this algorithm is able to change system parameters adaptively throughsensing environment noise. Theoretical analysis and simulation results show the algorithm has agood detection performance while sensing low SNR signal and it solves the problem of weak signaldetection effectively. And it overcomes the flexibility disadvantage of traditional stochasticresonance system.What’s more, against the problem that traditional stochastic resonance system can not be usedto detect high frequency signals, this paper proposes a cascading stochastic resonance spectrumsensing method which is based on the scale transformation. At first, the method uses the principle ofscale transformation to adjust the fixed parameters of stochastic resonance system in accordancewith a certain proportion. Meanwhile, amplify the mixed sampling signal of cognitive users so thatstochastic resonance system can be applicable for detecting high frequency primary user signal.Then, make two or more stochastic resonance systems become cascading to process the first stageoutput signal of stochastic resonance systems further. The method solves the problem of outputsignal high frequency noise glitch in traditional stochastic resonance system. Simulation resultsshow the algorithm can not only expand the application scope of traditional stochastic resonance system so that it can be used for detection high frequency signal but also improve the detectionprobability of the primary user signal.